We have recently shown that TolC and several TolC-dependent pumps reduce cellular stress in E. coli, presumably by expelling toxic cellular metabolites. In the course of these studies, we found that the CpxAR and BaeSR stress-response systems are activated in tolC mutants and in tolC+ cells bearing several mutations in certain TolC-dependent pump genes, including acrB. In an effort to identify these metabolites, we initiated metabolomic studies (with K. van de Wetering, Amsterdam) using liquid chromatography/mass spectroscopy (LC/MS). Our strategy was to look for compounds that accumulate inside the cells of efflux mutants but are deficient in the growth medium of the mutants because they cannot be exported from the cell. Analysis of the data showed that there are indeed significant differences between mutants and wild-type with respect to a modest number of low molecular weight compounds. These data suggest that the LC/MS method is potentially very useful for finding natural substrates of the pumps. One compound that we have tentatively identified as a substrate of the efflux pumps has the properties of 2,3-dihydroxybenzoyl-serine, a substance synthesized in the cell from 2,3-dihydroxybenzoate (2,3-DHB); 2,3-dihydroxybenzoyl-serine is trimerized to form enterobactin, a substance known to be excreted by the cell via TolC-dependent pumps. Others have recently shown that 2,3-DHB induces the Mar stress-response system. We have now shown that treatment of even wild-type cells with 2,3-DHB induces spy, a gene under the control of the CpxAR and BaeSR systems. Thus, 2,3-DHB may be an important signal for stress within the E. coli cell. We have also tested related compounds for their effects on stress-response systems and antibiotic-resistance. This work has now been completed and submitted for publication. No further work on this project is planned.
